Errors of memory in ATC
Shorrock, 2005 carried out research to determine the types of errors made in Air Traffic Control (ATC) primarily pertaining to errors in memory. The results were collected from the United Kingdom in two parts, by interviews with operational controllers and analyzing past incident reports. The results of this study will provide assistance in designing future Air Traffic Management (ATM) systems by considering ATC memory characteristics and integrating the evolving role of an air traffic controller effectively.
Results
Results showed that controllers make a combination of prospective and retrospective memory errors.
The result tables below are split into two categories:
- Internal error modes (IEM) - descriptions of how a cognitive function failed by forgetting, misrecalling or misplacing information.
- Psychological error mechanisms (PEM) - An explanation of the psychological underpinnings influencing an IEM.
Significance labels on the tables are a percentage of the total errors resulting in the following percentages.
- Rare <10%
- Occasional 10 > 25%
- Common > 25%
| Interview Data | Airprox Incident Report Data | ||
| Memory Error Category | Occurrence | Memory Error Category | Occurrence |
| Internal Error Modes | Internal error modes | ||
| Prospective memory failure | most common | Prospective memory failure | most common |
| Forget temporary information | common | Forget previous action | common |
| Forget stored information | occasional | Forget temporary information | occasional |
| Forget previous action | rare | Misrecall stored information | rare |
| Misrecall stored information | rare | Forget stored information | rare |
| Total Errors | 42 | Total Errors | 20 |
| Psychological Error Mechanisms | Psychological Error Mechanisms | ||
| Infrequency bias | most common | Distraction | most common |
| Memory capacity overload | occasional | Memory capacity overload | common |
| Distraction | rare | Preoccupation | occasional |
| Negative transfer | rare | ||
| Insufficient learning | rare | ||
| Total Errors | 21 | Total Errors | 7 |
The interview and incident report data did as anticipated show some inconsistencies between perceived everyday memory errors and memory errors that lead to an incident. For example forgetting previous actions was considered rare in interview results but was a common error causing incidents. This perception vs. factual evidence is valuable and can be used to counter errors more effectively through further education and training.
PEM results are of particular importance to individual controllers and managers in ATC alike. The mechanisms that resulted from the study show that PEM's like ineffectual training and distractions are a common factor in incidents.
One significant discrepancy in the results was the presence of the PEM factor infrequency bias (most common) in interviews but was not present in incident report findings. The results of this study may have identified a shortcoming in the current incident investigation analysis.
Methods
Research approach
This research was conducted in two parts using an exploratory method of interviewing air traffic controllers and secondary research by analysis of incident reports to identify errors of memory.
Sample
For part one a ready sample of twenty eight validated operational Area Controllers from the London Area and Terminal Control Centre were interviewed.
For part two 48 published 'Airprox' incident reports occurring in the London Area and Terminal Control Centre, Scottish Area Control Centre and Manchester Area Control Centre over 3 years, 1995-97 were analyzed.
Variables
Variables and limitations in this study were present. By collecting information from interviews the controllers were the variables. Each controller would have had different experience, professional opinions and willingness in providing information.
The interview subject was of a sensitive and personal nature and results may have been limited by interviewees censoring their responses because of this. Many of the errors identified were memory based and the recall of such events may have been forgotten or incorrectly described during interviews. Interviewers (two) may have made assumptions when partial or limited memory of an incident was available and this may have been inconsistent with the fellow interviewer. Overall the quality of the information derived from the interviews appears to be informative and honest.
The incident analysis results may have had methodological problems as the data had previously been published and were subject to the interpretations of the investigator and report author and based on the information known at the time of investigation. Some errors therefore may have been misinterpreted or misclassified.
Results had been determined for the satisfaction of an incident investigator and not for the objectives of this study therefore the incident report conclusions may not have identified additional (minor but not considered significant) errors of memory that were present.
Procedure
The controllers were interviewed individually by one of two interviewers in a closed office at the radar centre during a break from operational shift. The interviews lasted between 30 and 45 minutes and the controllers were assured that although the interviews were taped their responses would remain confidential.
Controllers were interviewed in a unstructured free flowing manner by being asked open questions initially and following up with probing questions to identify possible errors and occasionally asking closed questions to ensure accuracy in the interviewers understanding.
The transcript of the interviews were analyzed to identify when controllers mentioned any of five specific categories of errors (the interviews identified a number of wide ranging issues for this study only errors concerning memory were used).
Of the 48 Airprox reports only 27 of these incidents that involved errors of memory were used in this study.
Each Airprox incident report was analyzed using the TRACEr Classification to identify whether and error of memory had occurred and what error classification it was.
Data analysis
A descriptive analysis of results was undertaken with follow up discussion. This information can be found in the original article.
Generalization potential
This study, due to the limitations of sample size and breadth has limited generalization potential. The interviews and incident reports were collected only from the United Kingdom where it is likely that the systems, procedures and cultures that are present are likely to be different than many other parts of the world. By using only Area controllers from the UK for the interviews this limited the studies generalization to similar positions as controllers from other sectors use different systems, procedures and separation standards.
Both procedures used to collect results were vulnerable to second hand human interpretation of historic errors.
If a broader, larger sample was used consisting of all ATC positions and international air navigation service providers the generalization potential would have been considerably greater.
Want to know more?
Contributors to this page
Authors / Editors
| Other interesting sites |
 Knowledge |
 WikiofScience |
 AviationKnowledge |
 A4art |
 The Balanced Nutrition Index |
